Treubia 2003 33 (1) 35-42

PHYLOGENY OF Glyphodes Guenee (: : Spilomelinae) BASEDON NUCLEOTIDE SEQUENCE VARIATION IN A MITOCHONDRIAL CO I GENE: CONGRUENCE WITH MORPHOLOGICAL DATA

Hari Sutrisno Division of Zoology, Research Center for Biology, Indonesian Institute of Sciences, JI. Raya Bogor Km. 46, Cibinong, 16911 Indonesia E-mail: [email protected]

Abstract The phylogeny of Glyphodes Cuenee (14 species) and [our outgroup species (Feltia jaculifera, Metallarcha aureodiscalis, Talanga sexpunctalis and Agrioghrypta eurytusalis) was inferred from nucleotide sequence variation across a 686-bp region in the CO I gene. Over the entire 686-bp region, 19.9% sites were informative (3.35% in the 1"-, 0.29% in the 2''''- alld 16.32% in 3"'-codon position). The results also showed that the base composition of this region was high A+T biased (C= 0.258) and the averages of estimated sequence divergence in the comparisons between species within and between groups were 7.1% and 9.0%, respectively. In general, the phylogeny based 011 CO Igene by including all substitutions or any partial data set used in this study was not only able to recover almost the three monophyletic groups within Glyphodes as previously recovered by morphological phylogeny but also showed more clearly the relationships among them: Glyphodes group 2 was branched off first then followed by group 3 and 1.

Key words: CO I, Gbjphodes, morphology, phylogeny.

Introduction The Glyphodes Cuenee, 1854, is medium-sized (15-51 mm) belongs to the subfamily Spilomelinae (Common, 1990; Shaffer et al., 1996). This is diverse and widespread in tropical regions with some species penetrating into subtropical and warm temperate areas (Common, 1990;Robinson et al., 1994).The genus consists of 120 species throughout the world, and about 25 and 17 species have been recorded in the Southeast Asia and Australia, respectively (Robinson et al., 1994; Shaffer et al., 1996). As other pyraloids in the Indo-Australian regions, this genus is also poorly defined and has been suggested that it is seriously in need of revision since the current generic classification of this genus was based on superficial similarities only (Robinson et al., 1994;Sutrisno & Horak, 2003). However, the comprehensive study on the or phylogeny to assess the monophyly and the relationships among them has never been conducted except for a cladistic analysis based on the Australian Glyphodes and its allied genera using morphological data (Sutrisno, 2002). The cladistic analysis showed that the genus Glyphodes falls into three monophyletic groups (Glyphodes group I, 2 and 3); each of them can be diagnosed by the adult morphology. In general, the result showed that morphological characters used 36 Sutrisno: Phylogeny of Glyphodes Guenee (Lepidoptera: Crambidae: Spilomelinae)

in the previous study were able to infer the phylogenetic relationships of this genus. However, the evolutionary change of morphological characters in this genus is extremely complicated that this approach did not produce a clear-cut picture of evolutionary history; i.e., for the relationships among groups within Glyphodes. In addition, the relationships among closely-related species within Glyphodes group 3 seem very difficult to be resolved. Therefore, another possible approach such as molecular phylogeny that has been proved to give more powerful resolution is really needed. Mitochondrial genes have been used intensively to infer the phylogenetic relationships among groups of as have been repeatedly reported by numerous authors (reviewed in Simon et al., 1994). CO I gene is one of the most common to be considered infering the relationships among closely-related species in several groups of Lepidoptera, as individual gene or to be combined with other genes (Andrew, 1994; Sperling & Hickey, 1994; Caterino & Sperling, 1999; Landry et al., 1999; and Blum et al., 2003). Therefore, in this issue I used mitochondrial CO I gene to reassess the monophyly of Glyphodes and to estimate the relationships among species within this genus.

Material and Methods DNA extraction and Sequencing A total of 17 species representing the four genera (Glyphodes, Talanga, Agrioglypta and Metal/archa) from various localities in Indonesia and Australia were collected using a light trap and preserved in absolute alcohol (Table 1). The DNA extraction method and the PCR amplification conditions used in this study were similar with the procedures described in my previous study (Sutrisno, 2003). I used Mt 0-4 and Mt 0-9 primers to amplify CO I gene for a total of 686-bp and the complete sequences were Mt 0-4: 5'- TACAATTTATCGCCTAAACT TCAG CC-3', and Mt 0-9: 5'-CCCGGTAAAATT AAAATATAAACTTC-3' (http://www.biotech.ubc.ca/services/naps/primers.html). Sequencing was performed using an AB!PRISMDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer) on AB! PRISM model 310 Genetic Analyzer (PE Applied Biosystems). Phylogenetic Analysis For the phylogenetic analysis, I used Maximum-Parsimony method using an exhaustive search with different data sets: all substitutions; transversions only; weighting scheme 2:1; ntl + nt3; and nt3 only and Neighbor Joining method using K80 model (Kimura, 1980) in PAUP 4.0b4a (Swofford, 2001); and Bootstrap test (Felsenstein, 1985) to evaluate the statistical confidence of the MP and NJ trees. Four species, Agrioglypta eurytusalis (Walker), Talanga sexpunctalis Moore, Metallarcha aureodiscalis Meyrick and Feltia jaculifera Meyrick (Noctuidae) were included as outgroup taxa in the present analysis. The sequence of CO I-CO II of the last species is available from the gene bank with the accession number U60990. Treubia 2003 33 (1) 37

Table 1. Species examined for the molecular study

Species Collection locality Date of Collection

Agrioglypta eurytusalis (Walker) G. Pangrango, Java September 2000 Glyphodes G. actorionalis (Walker) Sorong, Papua February, 2002 G. apiospila (Turner) Sorong, Papua February, 2002 G. bicolor (Swainson) Sorong, Papua February, 2002 G. bivitralis Cuenee Patunuang, Sulawesi April, 2001 G. caesalis (Walker) Menado, Sulawesi March,2001 G. conjuncialis Walker Sorong, Papua February, 2002 G. cosrnarcha Meyrick Patunuang, Sulawesi April,2001 G. doleschalii Lederer Sorong,Papua January, 2002 G. jlauizonalis Hampson Sorong, Papua January, 2002 G. margaritaria (CIerck) Patunuang, Sulawesi April, 2001 G. multilinealis Kenrick Bantimurung, Sulawesi April, 2001 G. onychinalis Cuenee Bucasia, Queensland April, 2001 G. puioerulentalis Hampson Sukabumi, Java February, 2002 G. stolalis Cuenee G. Halimun NP, Java March,2002 Metallarcha aureodiscalis Meyrick Bucasia, Queensland April,2001 Talanga sexpuncialis Moore Patunuang, Sulawesi April, 2001

Table 2. Percent variable sites across 14species of Glyphodes and four outgroup species

Total lst-codon 2nd-codon 3rtl-codon

Constant (%) 67.9 27.25 31.77 8.89 Uninformative (%) 12.1 2.76 1.31 8.01 Informative (%) 19.9 3.35 0.29 16.32 Number of sites 686 229 229 228

Results Base compositions Sequences of 14 species of the Glyphodes, and four species outgroup were aligned with no evidence of insertion and deletion (aligned sequences are available from the author on request). Over the entire 686-bp region, 67.9 % of the nucleotide positions were constant, 12.1 % were uninformative (i.e., any variants were found in single sequence), and 19.9% were informative (Table2). In addition, the most informative 38 Sutrisno: Phylogeny of G/yphodes Guenee (Lepidoptera: Crambidae: Spilomelinae)

sites for maximum parsimony analysis lied in the third-codon positions and the least informative site was in the second-codon positions. Table 3 shows the proportion of nucleic acids and bias in CO I. The results showed that the base compositions were strongly biased (C= 0.258) with the A+T proportions ranged from 69.82% to 72.01%. Interspecific variations in base compositions CO I were very low for a total of nucleotides (the Chi-square test, X2= 10.684, df= 51, P=1.00). In addition, the averages of estimated sequence divergence in the comparisons between species within and between groups were 7.1%and 9.0%, respectively.

Table 3. Proportion of each nucleotide and bias in CO I

Codon position

l'l-codon 2nd-codon 3rd-codon Mean

A 0.316 0.170 0.485 0.324 C 0.139 0.249 0.042 0.142 G 0.264 0.169 0.097 0.175 T 0.279 0.410 0.462 0.384 Bias 0.258

Bias in base composition is calculated as

4 C = (X)2J, - 0251

/::1

Where C is the composition bias and ci is the frequency of its base.

Phylogenetic analyses An exhaustive search in maximum-parsimony analysis using all nucleotides substitutions resulted in two equally MP trees (Fig.l). The MP trees showed that the three groups within the Glyphodes as previously recognized in the morphological study (Sutrisno, 2002) were recovered. However, the relationships among species within Glyphodes group 2 were less-resolved: G. onychinalis was evolved independently. When the weighting scheme 2:1,or transversional substitutions only or the combination between 1"_and 3rd-codon (nt1+nt3) or 3rd-codon (nt3) only were included in the analysis, the relationships among five species of Glyphodes group 2 was also poorly-resolved, except for the relationships between G.flavizonalis and G. apiospila (Figs. 2-5). The descriptive tree statistics for parsimony analyses are presented in Table 4. The NJ tree based on all substitutions revealed that each group was found to be monophyletic (Fig. 6) and the relationships among them were well-resolved: Glyphodes group 2 was branched off first then followed by group 3 and 1. In addition, the relationships among the three species of Glyphodes group 3 were also well-resolved in both MP and NJ analyses: G. doleschalii lied in the basal group, G. actorionalis was the sister group of G. bivitralis. Treubia 2003 33 (1) 39

F.jaculifera Fjaculifera M. aureodiscalis M. aureodiscalis 5 A eurytusalis 58 A eurutusalis T sexpunctalis T sexpunclalis 100 50 Gflavizonalis ] 55 Gflavizonalis ] G apiospila ~ 3 G apiospila ~ I------G onychinalis .g ,------+--G conjunctalis -ii 1------1iG conjunctalis 2 G bicolor 20 G bicolor 19f------G onychinalis 51 G caesalis ] G. pulverulen talis g ;:~~a:;~lentaliS] Cl ,----G margaritaria ~ ,----G margaritaria g 34 ~~~~li~cha .g 36 G ccsmar cha '" g: G stalalis G multilinealis G multilinealis 97 . G actorionalis ] Cl 97 G actarianalis ] ClCl 1 85 G bivitralis g 8 2 G bivitralis -ii G doleschalii '" G dolescholii 3 3

Fjaculifera F [acul ifera M. aureodiscalis M aureodiscalis 51 A eurytusalis 69 A eurytusalis T sexpunclalis T sexpunctalis 41 Gflavizonalis ] lOO G.flavizonalis ] G apiospila Cl 62 G. apiospila ",QS1 f------G conjunctalis g f------G.onychinalis _ f------G bicalor '" G bicolor f------G onychinalis 2 G conjunctalis 2

25 57 ~:~~a::~lentalisl g: g: ;:~::~lentaliS] Cl f----G margaritaria ~ r----G margaritaria g G cosmarcha -ii 44 35 G cosmarcha '" G stolalis G. 5tolalis 1 G multilinealis G. multilinealis 9 G actorionalis ] Cl G actorionalis 4 94 3 66 G bivitralis g 88 G bivitralis ] G doleschalii -o G. dolesdialii

F jaculifera ,------Fjacu/ifora M. aureodiscalis M aureodisealis 48 A eurytusalis A. eurytusalis T sexpunctalis T sexpunetalis 57 Gflavizonalis ] G. flavizonalis ] G apiospila Cl r: 0. apiospi/a ~ f------G onychinalis g 55 0. bicolor -e G hi calor -e "----G. eonjunetalis 2 G conjun ctalis 2 G onychinalis G caesalis ] 17 51 G caesalis ] G puluerulen talis Cl G pulverulentalis Q I-----G margaritaria g G margarttarta ~ G cosmar chu '" G eosmareha 40 G stolalis 36 73 G stolalis G multilinealis G. multilinealis 98 G. actorionalis ] Q G actorionalis ] "'~ 5 70 G biuitralis :: 6 G. bivitralis 0 0. doleschalit -e G doleschalii 3 0.05 substitutions/site

Figs. 1-6. MP and NJ trees, 1, Strict consensus of the two MP trees based on all substitutions; 2, Strict consensus of the two MP trees based on weighting scheme 2:1;3, Strict consensus of the six MP trees based on nt2+nt3; 4, Strict consensus of the two MP trees based on ntl +nt3; 5,Strict consensus of the six MP trees based on nt3 only; 6, NJ tree based on all substitutions. 40 Sutrisno: Phylogeny of G/yphodes Guenee (Lepidoptera: Crambidae: Spilomelinae)

Table 4. Descriptive tree statistics for parsimony analyses

Cl RI Steps No of Trees Tree Topology

Nucleotide All substitutions 0.388 0.423 546 2 Figure 1 Weighting scheme 2:1 0.378 0.424 869 2 Figure 2 nt2 + nt3 0.388 0.415 456 6 Figure 3 ntl + nt3 0.368 0.412 534 2 Figure 4 nt3 0.385 0.415 444 6 Figure 5

Note: ntl= 1st-codon position; nt3= 3'd-codon position

Discussion It has been reported that the base composition in mitochondrial genomes in general is high A+T biased (reviewed in Simon et al., 1994). Partial sequences of CO I from 21 species of Drosophila also showed high A+T proportion ranged from 68.9% to 71.4% (Goto & Kirnura 2001). In Lepidoptera, high A+Tcontents have been found also in CO I of Choristoneura (Sperling & Hickey, 1994) and Hemileuca (Rubinoff & Sperling, 2002) which ranged from 62% to74%. The A+T proportion in the present study (69.82- 72.01%) was comparable with those found in other Lepidoptera and Drosophila. In addition, the bias in base compositions was found to be the greatest at third-base position. This is perhaps because the first- and second-codon position are more constrained by the amino acid composition of the encoded protein (Wolstenholme & Clary, 1985; Brown et al., 1994). It has been suggested that mitochondrial genes are the best candidate for resolving relationships among closely-related taxa because they do not recombine and usually occur in only one haplotype per individual (Brower & DeSalle, 1994) and the mitochondrial gene CO I as individual gene, or combined with CO II has been reported to be able to resolve the relationships in Argyrotaenia franciscana species group of Tortricidae (Landry et al., 1999), Choristoneura fumiferana species group (Sperling & Hickey, 1994) and genus Papilio (Caterino & Sperling, 1999). In addition, the combination of the CO I and EF-lti increased resolution and supports most of the phylogenetic relationships suggested by separate analysis of each gene in the genus Hemileuca (Rubinoff & Sperling, 2002). The results of this study also showed that the MP and NJ analyses based on CO I resolved the relationships among species within Glyphodes eventhough with low bootstrap supports for each group except for the group 3. There is no doubt that including more characters by adding a longer sequence of CO I or other mitochondrial genes may increase the resolution and supports for the relationships suggested in the present study. Treubia 2003 33 (1) 41

It is not surprising that group 3 is always has a strong bootstrap support since the morphological study also showed that this group was supported by several good synapomorphies. In addition, the analysis based on CO I also resulted better resolutions on the relationships among the three species of this group, which hardly possible to assess using morphological analysis since their male and female genitalia were quitle similar and the evolutionary changes of those characters were difficult to be 'distinguished and scored as informative characters (Sutrisno, 2002). In general, all the findings in the present study suggest that phylogeny of Glyphodes inferred from a mitochondrial COI gene was congruent with the morphological data: Glyphodes was divided into three groups; Glyphodes group 1 was branched off first then followed by group 3 and 1. Further studies are needed to be done by including more species and other mitochondrial genes in order to test the validity of the relationships proposed here.

Acknowledgments Grateful thanks due to Dr. M. Horak (CSIRO), Canberra and Mr. Ken ]. Sandery for collecting and sending the specimens from Australia. My thanks also go to Ms Purwaningrum and Dr. D. Satriatmi (Fac. of Agriculture, IPB) for rearing G. pulverulentalis, E. Cholik, Darmawan K Sofyan (Museum Zoologicum Bogoriense) for helping me in collecting specimens at G. Halimun NP.

Literature Cited Andrew, V. Z.B. 1994. Phylogeny ofHeliconius ButterfliesInferred from Mitochondrial DNA Sequences (Lepidoptera: Nymphalidae). Mol. Phylogenet. Evol., 3(2):159-174. Blum, J. M., Bermingham, E., and Dasmahapatra, K. 2003. A molecular phylogeny of the neotropical butterflies genus Anartia (Lepidoptera: Nymphalidae). Mol. Phylogenet. Evol., 26(1),46-55. Brower, A. V. Z. & DeSalle, K 1994. Practical and Theoretical Consideration for Choice of a DNA Sequence Region in Insect Molecular Systematics, with a Short Review of Published Studies Using Nuclear Gene Regions. Ann. Entomol. Soc., 87(6):703-716. Brown, J. M., Pellmyr, 0., Thompson, J. N., and Horrison., K, G. 1994. Phylogenyof Greya (Lepidoptera: Prodoxidae), based on nucleotide sequence variation in mitochondrial cytochrome oxidase I and II: congruence with morphological data. Mol. BioI. Evol., 11 (1):128-141. Caterino, M. S. & Sperling, F. A. 1999. Papilio Phylogeny Based on Mitochondrial Cytochrome Oxidase I and II genes. Mol. Phylogenet. Evol., 11(1), 122-137. Common, I. F. C. 1990. of Australia. Melbourne University Press, Carlton. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the boots trap. Evolution, 39: 783-791. 42 Sutrisno: Phylogeny of G/yphodes Guenee (Lepidoptera: Crambidae: Spilomelinae)

Goto, S.G., & Kimura, M.T. 2001. Phylogenetic Utility of Mitochondrial CO I and Nuclear Gpdh genes in Drosophila. Mol. Phylogenet. Evol., 18 (3): 404-422. Irwin, D. M., Kocher, T. D., and Wilson, A C 1991. Evolution of the Cytochrome b gene of Mammals. Jour. Mol. Evol., 32: 128-144. Kimura, M. 1980. A simple method for estimating evolutionary rate of base substitution through comparative studies in nucleotide sequences. Jour.Mol. Evol., 16:111-120. Landry, B., Powell, J. A, and SperIing, F. A H. 1999. Systematics of the Argtjrotaenia franciscana (Lepidoptera: Tortricidae) species group: Evidence from Mitochondrial DNA Systematics, 92(1): 40-46. Robinson, G. S., Tuck, K. R., and Shaffer, M. 1994. Field Guide to the Smaller Moths of Southeast Asia. Malysian Nature Society, Kuala Lumpur. Rubinoff, D. & Sperling, F.AH. 2002. Evolution of ecological traits and wing morphology in Hemileuca (Saturniidae) based on a two-gene phylogeny. Mol. Phylogenet. Evol., 25(1): 70-86. Shaffer, M. A, Nielsen, E. S., & Horak, M. 1996. Pyraloidea. In: Nielsen, E. S., Edwards, E. S. & Rangsi, T. V. (ed). Checklist of the Lepidoptera of Australia. CSIRO Australia, 164-199. Simon, C, Frati, F., Beckenbach, AT., Crespi, B.,Liu, H., and Flook, P. 1994. Evolution, Weighting, and Phylogenetic utility of Mitochondrial Gene Sequences and a Compilation of conserved Polymerase Chain Reaction Primers. Ann. Entomol. Soc., 87(6): 651-701. Sperling, F.AH., & Hickey, D.A 1994. Mitochondrial DNA Sequences Variation in the Spruce Budworm Species Complex (Choristoneura: Lepidoptera). Mol. Bioi. Evol., 1(4),656-665. Sutrisno, H. 2002. Cladistic analysis of the Australian Glyphodes and allied genera (Lepidoptera: Crambidae; Spilomelinae). Entomol. Sci., 5(4): 457-467. Sutrisno, H. & Horak, M. 2003. Revision of the Australian species of Hyalobathra Myerick (Lepidoptera: Pyraloidea: Cerambidae: ) based on adult morphology with description of a new species. Aust. Jour. Eniomol., 42: 233-248. Sutrisno, H. 2003. Phylogeny of the two closely-related genera, Agn'oglypta Myerick and TalangaMoore (Lepidoptera: Crambidae; Spilomelinae)based on nucleotide sequence variation in mitochondrial cutochrome oxidaseII and morphology. (Submitted). Swofford, D.L. 2001. PA UP*. Phylogenetic Analysis Using Parsimony (* and Other Methods). Version 4.0b10 for 32-bit Microsoft Windows. Sinauer Associates, Sunderland, Massachusetts. Wolstenholme, D.R. & Clary, D.O. 1985. Sequence evolution of Drosophila mitochondrial DNA Genetics, 109: 725-744.

Accepted November 2003